Re: st: Re: ivprobit not converge

[Maarten buis <maartenbuis@yahoo.co.uk>]

--- On Fri, 20/8/10, xueliansharon wrote:
> But when I run the ivprobit model many times with the same
> codes and data set, it converges sometimes. And the
> parameter estimates are similar in those converged cases.
> So are these coefficients reliable?

I would be reluctant. The maximization algorithms used by 
Stata are pretty good, so the fact that -ivprobit- sometimes 
does not converge is for me reason enough to suspect that 
your model tries to extract more information than the data 
contains. 

So I will stick to my earlier recomendation, which is to 
start with a very simple model, and add complexities step 
by step, till you see this non-converging behaviour, and
see if you understand why the complexity you added last 
causes trouble. Once you understand why this problem 
occurs, you can make an informed decision whether or not 
you need to do something about it, whether or not you can
do something about it, and what that something might be.

Hope this helps,
Maarten

--------------------------
Maarten L. Buis
Institut fuer Soziologie
Universitaet Tuebingen
Wilhelmstrasse 36
72074 Tuebingen
Germany

http://www.maartenbuis.nl
--------------------------


      

*
*   For searches and help try:
*   http://www.stata.com/help.cgi?search
*   http://www.stata.com/support/statalist/faq
*   http://www.ats.ucla.edu/stat/stata/